1 Insights into the evolution of mesophily from the bacterial phylum Thermotogae
1 2
Stephen M. J. Pollo, Olga Zhaxybayeva, Camilla L. Nesbø 3
4
Stephen M. J. Pollo 5
Department of Biological Sciences, 11455 Saskatchewan Drive, University of Alberta, Edmonton, 6
Alberta, Canada, T6G 2E9. pollo@ualberta.ca 7
Olga Zhaxybayeva 8
Department of Biological Sciences and Department of Computer Science, Dartmouth College, 78 9
College Street, Hanover, NH, 03755, U.S.A. olga.zhaxybayeva@dartmouth.edu 10
Camilla L. Nesbø 11
Department of Biological Sciences, 11455 Saskatchewan Drive, University of Alberta, Edmonton, 12
Alberta, Canada, T6G 2E9 and Centre for Ecological and Evolutionary Synthesis (CEES), 13
Department of Biology, University of Oslo, P.O. Box 1066 Blindern, N-0316 Oslo, 14
Norway. nesbo@ualberta.ca and c.l.nesbo@bio.uio.no 15
16
Corresponding Author:
17
Camilla L. Nesbø 18
Department of Biological Sciences, CW 405 Biological Sciences Bldg., 11455 Saskatchewan 19
Drive , University of Alberta, Edmonton, Alberta, Canada, T6G 2E9. Telephone: (01) 780-492- 20
8956 nesbo@ualberta.ca and c.l.nesbo@bio.uio.no 21
Running Title: Temperature adaptation in Thermotogae 22
2 Abstract: Thermophiles are extremophiles that grow optimally at temperatures > 45°C. In order 23
to survive and maintain function of their biological molecules, they have a suite of characteristics 24
not found in organisms that grow at moderate temperature (mesophiles). At the cellular level, 25
thermophiles have mechanisms for maintaining their membranes, nucleic acids and other cellular 26
structures. At the protein level, each of their proteins remains stable and retains activity at 27
temperatures that would denature their mesophilic homologs. Conversely, cellular structures and 28
proteins from thermophiles may not function optimally at moderate temperatures. These 29
differences between thermophiles and mesophiles presumably present a barrier to evolutionary 30
transitioning between the two lifestyles. Therefore, studying closely related thermophiles and 31
mesophiles can help us determine how such lifestyle transitions may happen. The bacterial 32
phylum Thermotogae contains hyperthermophiles, thermophiles, mesophiles and organisms with 33
temperature ranges wide enough to span both thermophilic and mesophilic temperatures.
34
Genomic, proteomic and physiological differences noted between other bacterial thermophiles 35
and mesophiles are evident within the Thermotogae. We argue that the Thermotogae is an ideal 36
group of organisms for understanding both the response to fluctuating temperature as well as 37
long-term evolutionary adaptation to a different growth temperature range.
38 39
Key Words: lateral gene transfer, Kosmotoga, Mesotoga, thermostability, stress response 40
41
3 Introduction
42
Extremophiles are organisms that thrive under extreme environmental conditions unsuitable for 43
survival of most other organisms. As such, they are of great interest for delineating the limits of 44
conditions that permit life’s existence, a key insight needed to advance efforts in the search for 45
life on Earth and other planets (Pikuta et al. 2007; Rothschild and Mancinelli 2001). Additionally, 46
due to their intrinsically “extreme” nature, these organisms are also desirable sources of enzymes 47
and other biomolecules that function under conditions that render other organisms and their 48
enzymes inactive. Such biomolecules may have a wide range of biotechnological and industrial 49
applications from clean energy to bioremediation and carbon sequestration.
50
When examining temperature as a parameter that can either permit or exclude life, there 51
are mesophiles, the organisms that grow optimally at moderate temperatures, and two types of 52
extremophiles: psychrophiles, which grow optimally at temperatures below 15°C, and 53
thermophiles, which grow optimally at temperatures above 45°C (Kimura et al. 2013). Within 54
thermophiles, organisms growing optimally at > 80°C are commonly referred to as 55
hyperthermophiles. Thermophiles are of particular interest due to their ability to withstand the 56
denaturing effect of higher temperatures on biological molecules such as proteins and DNA (Li et 57
al. 2005).
58
The phylogenetic position of the hyperthermophile-containing bacterial lineages 59
Thermotogae, Thermodesulfobacteria and Aquificae at, or close to, the base of the 16S rRNA tree 60
of life (Fig. 1), has been used as support for the hypothesis that the ancestor of the bacterial 61
domain was a hyperthermophile (Achenbach-Richter et al. 1987). Similarly, thermophilic 62
Archaea are also found at the base of the Archaeal domain (Fig. 1). Together with the proposed 63
high temperature conditions of early Earth this led to the hypothesis that the last universal 64
4 common ancestor (LUCA) was a hyperthermophile (Pace 1991). A (hyper)thermophilic LUCA is 65
also supported by experimental evidence from resurrection of ancestral nucleoside diphosphate 66
kinases and characterizing their properties (Akanuma et al. 2013). Other lines of evidence, 67
however, suggest that the LUCA may have been either a mesophile or a thermophile growing 68
optimally below 80°C (Boussau et al. 2008; Brochier-Armanet and Forterre 2006). Whether the 69
LUCA lived at the time of life's origin or much later remains debatable as well (Zhaxybayeva and 70
Gogarten 2004).
71
Regardless of the optimal growth temperature of the LUCA, the ancestors of present day 72
bacterial and archaeal lineages have had to modify their cellular structures and protein 73
compositions to transition between mesophilic and thermophilic lifestyles (Boussau et al. 2008).
74
Given the distribution of mesophiles and thermophiles on the Tree of Life (Fig. 1), we infer that 75
such transitions likely happened independently multiple times. This same inference has been 76
made based on multivariate analyses of the amino acid compositions of 279 prokaryotes (Puigbò 77
et al. 2008) and from the different mechanisms of DNA supercoiling and the phylogeny of the 78
involved genes (López-García 1999). This conjecture is also supported by reconstruction and 79
synthesis of ancestral versions of enzymes and examining the optimal temperature at which they 80
function. For example, examination of LeuB enzymes (3-isopropylmalate dehydrogenase) in the 81
Bacillus genus suggests multiple transitions between thermophilic and mesophilic temperature 82
optima when going forward in evolutionary time from the Bacillus ancestor (Hobbs et al. 2012).
83
Therefore, thermophily has been lost and gained throughout the evolutionary history of the genus 84
Bacillus. Similarly, analysis of extant and reconstructed ancestral myo-inositol-3-phosphate 85
synthase enzymes from Thermotoga and Thermococcales suggests higher optimal growth 86
temperatures of the ancestors (Butzin et al. 2013), indicating fluctuations of the tolerated 87
5 temperature ranges of these organisms throughout their evolutionary history. Together these 88
studies imply that temperature adaptations may not be too difficult, and the growth temperature 89
range may change rapidly and frequently in many lineages.
90
Temperature adaptation can be defined either as a response of an individual cell to 91
changes in temperature, or as an evolutionary adaptation of an organismal lineage (such as 92
‘species’) to growth within a certain temperature range. To distinguish between the two, we will 93
refer to temperature response for the former and temperature adaptation for the latter. These two 94
phenomena are related, as selection acting on temperature responses may eventually lead to 95
temperature adaptations. In this review we focus on organismal responses and lineage adaptations 96
to moderate and high temperatures. For a review of adaptation to very low growth temperatures 97
see Siddiqui et al. (2013). Specifically, we will discuss properties of thermophiles, and how these 98
properties may relate to a transition between thermophily and mesophily, with a particular 99
emphasis on the bacterial phylum Thermotogae.
100 101
The Thermotogae 102
Bacteria belonging to the Thermotogae phylum were first isolated by Karl Stetter and colleagues 103
in 1986 from geothermally heated sea floors (Huber et al. 1986). Their name derives from the 104
unique outer sheath-like structure that balloons over each end of the cell, known as the “toga”
105
(Fig. 2) (Huber et al. 1986). There are 12 described genera in this phylum, most of which are 106
thermophiles (Fig. 3). In the accepted taxonomy, these genera are all grouped in a single order, 107
Thermotogales, and one family, Thermotogaceae. However, a reclassification of these bacteria 108
into separate orders is overdue, and a division into three orders and four families has been 109
6 recently proposed (Bhandari and Gupta (2014); Fig. 3). While the new classification is based on 110
conserved indels, it is consistent with the 16S rRNA phylogeny (Fig. 3).
111
Thermotogae are anaerobes and organotrophs, capable of growing on a wide range of 112
complex substrates (Conners et al. 2006). They are found in hot ecosystems all over the world 113
including thermal springs, hydrothermal vents, and petroleum reservoirs (Huber and Hannig 114
2006; Ollivier and Cayol 2005), with some members growing at temperatures up to 90°C.
115
Although it was long thought that the Thermotogae only harbored thermophiles and 116
hyperthermophiles (11 of 12 genera are entirely composed of thermophiles or 117
hyperthermophiles) (Fig. 3), mesophilic Thermotogae from the genus Mesotoga have recently 118
been detected and isolated from cool hydrocarbon-impacted sites such as oil reservoirs and 119
polluted sediments (Ben Hania et al. 2011; Ben Hania et al. 2013; Nesbø et al. 2006b; Nesbø et 120
al. 2010; Nesbø et al. 2012). Interestingly, the closest relative of Mesotoga, Kosmotoga olearia, 121
has an unusually wide growth temperature range, which may have been important in Mesotoga’s 122
adaptation to low temperature (DiPippo et al. 2009; Nesbø et al. 2012).
123
As of May 2015, over 80 completed and ongoing Thermotogae genome projects 124
comprise 10 of the 12 described Thermotogae genera, with no genome projects for Geotoga nor 125
Oceanotoga (Benson et al. 2014; Reddy et al. 2014). The maximum divergence in the 16S rRNA 126
genes of these cultivated Thermotogae is ~25%, similar to what is observed for other bacterial 127
phyla (Konstantinidis and Tiedje 2005). For protein coding genes pairwise average amino acid 128
identity (AAI; Konstantinidis and Tiedje 2005) between genera ranges from 45 to 69% (average 129
49%). Phylogenetic analysis of environmental 16S rRNA gene sequences shows several novel 130
Thermotogae lineages without any cultivated members, and based on the nucleotide identity they 131
would be classified as new genera (Nesbø et al. 2010). Thus, as with most microbial lineages, 132
7 there is a large unknown diversity of Thermotogae. At least four of these new lineages have only 133
been detected in low temperature environments (as low as 9.5°C), suggesting that Thermotogae 134
might be common in mesothermic environments. Interestingly, on the phylogenetic tree these 135
likely mesophilic lineages fall within multiple thermophilic clades (Nesbø et al. 2010), 136
suggesting several independent adaptations to lower temperatures.
137
With mesophilic Thermotogae only recently discovered, the functional characterization of 138
this phylum has focused on thermophiles, mainly the hyperthermophilic organisms Thermotoga 139
maritima and Thermotoga neapolitana. Protein crystal structures have also been experimentally 140
determined for a large portion of the T. maritima proteome (DiDonato et al. 2004; Lesley et al.
141
2002), and the protein structures of its central metabolic networks were modeled by Zhang et al.
142
(2009). Complimented with models of high temperature hydrogen and sulfur metabolism 143
(Cappelletti et al. 2014; Schut et al. 2012), this wealth of functional information makes the 144
Thermotogae a promising microbial lineage for industrial and biotechnological applications. For 145
example, most Thermotogae produce hydrogen that may be harvested (e.g., Nguyen et al. (2008) 146
and Maru et al. (2012)). The hydrogen production of T. maritima can be boosted via metabolic 147
engineering, as was demonstrated by an in silico re-design of its metabolism (Nogales et al. 2012).
148
Additionally, while the degradation of sugars by many Thermotogae results in the production of 149
CO2 and acetate, T. neapolitana has been shown to convert these by-products to lactic acid when 150
grown in a CO2 atmosphere, a process suggested to have potential in carbon capture (D'Ippolito et 151
al. 2014).
152
Carbohydrate utilization by T. maritima has been examined by studying the substrate 153
specificities and affinities of its carbohydrate transporters (Boucher and Noll 2011; Cuneo et al.
154
2009; Ghimire-Rijal et al. 2014; Nanavati et al. 2005; Nanavati et al. 2006) and their 155
8 transcriptional regulation in response to growth on different saccharides (Frock et al. 2012).
156
Information about substrate specificities, enzymatic activities and catalytic mechanisms of many 157
of T. maritima’s glycoside hydrolases are also available (Arti et al. 2012; Comfort et al. 2007;
158
Kleine and Liebl 2006), which has been used, for instance, to engineer an alpha-galactosidase 159
from T. maritima into an efficient alpha-galactosynthase (Cobucci-Ponzano et al. 2011). The 160
transcriptional regulation of glycoside hydrolases and other carbohydrate metabolism-related 161
genes in response to growth on various carbohydrates highlights the differences in carbohydrate 162
utilization, even between closely related Thermotogae lineages (Chhabra et al. 2002; Chhabra et 163
al. 2003; Frock et al. 2012). Moreover, interconnections exist between sugar regulons in T.
164
maritima’s carbohydrate utilization network, suggestingcoordinated regulatory responses to 165
particular types of complex carbohydrates (Rodionov et al. 2013). This rich knowledge base will 166
be very useful in comparative studies of thermophilic and mesophilic Thermotogae lineages and, 167
ultimately, will lead to understanding processes leading to shifts in an organism’s growth 168
temperature range.
169 170
General cellular adaptations to thermophily 171
Regardless of whether cells are responding to transient temperature increases within their growth 172
range or evolving to an alternate growth range, changes in temperature require major 173
modifications across the cell to optimize cell function and growth. The following sections discuss 174
some of these temperature responses and adaptations in microbial cells.
175 176
The effect of temperature on cellular membranes: maintaining a fluid envelope 177
9 The cell membrane is critical to cell function since it maintains and separates the interior cell 178
environment from the exterior environment. In order to serve its function, a lipid membrane must 179
be impermeable to most solutes and maintain a liquid crystalline phase, even under stress (de 180
Mendoza 2014). As the temperature increases, membrane integrity and impermeability become 181
compromised, which eventually results in cell death (Chang 1994). Therefore, thermophiles must 182
maintain their membranes under conditions that could inactivate those of mesophiles. Bacteria 183
and Archaea handle this challenge differently due to the dissimilar structures of their membrane 184
lipids (reviewed in detail by Oger and Cario (2013), Koga and Morii (2005), Koga (2012), and 185
Mansilla et al. (2004)). We will only focus on bacterial lipids here. For a review on archaeal lipids 186
see Oger and Cario (2013).
187
Bacterial polar membrane lipids consist mainly of straight-chain fatty acids that are bound 188
to the polar head group predominantly by ester linkages (Koga and Morii 2005). Bacteria respond 189
to various temperatures by altering the composition (length, degree of branching and degree of 190
unsaturation) of their fatty acid chains to maintain membrane fluidity (Mansilla et al. 2004; Zhang 191
and Rock 2008). The types of fatty acids bacteria are able to produce will therefore influence the 192
temperature range within which they can grow. For example, hyperthermophilic Thermotogae 193
have unusual membrane-spanning diabolic fatty acids in their membrane, which are thought to be 194
an adaptation to high temperature growth (Carballeira et al. 1997; Damsté et al. 2007). In 195
agreement with this hypothesis, these diabolic fatty acids are not found in the membranes of the 196
mesophilic Mesotoga prima (Nesbø et al. 2012). Moreover, M. prima (grown at 35°C) contained 197
branched, mono-unsaturated and saturated fatty acids, while K. olearia (grown at 55°C) contained 198
only saturated fatty acids (Nesbø et al. 2012). Fatty acid composition is also part of the immediate 199
cold-shock response with genes involved in production of, for instance, branched fatty-acids 200
10 being up-regulated in the thermophile Thermoanaerobacter tengcongensis when grown at sub- 201
optimal temperatures (Liu et al. 2014). Increase of branched fatty acids is a common response to 202
lower temperatures (Suutari and Laakso 1994), and in Listeria monocytogenes this is due to 203
temperature-dependent substrate selectivity of FabH, the enzyme responsible for the first 204
condensation reaction in fatty acid biosynthesis (Singh et al. 2009). Interestingly, in Bacillus a 205
transmembrane two-component response regulator, which controls the desaturase that introduces 206
double bonds in preexisting fatty acids, senses changes in membrane fluidity and not the actual 207
temperature changes (de Mendoza 2014).
208
In addition to the lipid structure of cell membranes, integral membrane proteins affect the 209
temperature tolerance of an organism (Thompkins et al. 2008). Therefore, while the lipid 210
composition of the membrane is crucial for its function, integral membrane proteins may also 211
play a significant role, particularly with respect to the temperature limit of an organism’s growth 212
range. For example, mutations of integral membrane proteins of the DedA family cause 213
temperature sensitivity and cell division defects in Escherichia coli (Thompkins et al. 2008).
214
Interestingly, proteins from the DedA family have been shown to be essential in at least two 215
bacterial species (E. coli and Borrelia burgdorferi), but their homologs are not detected in 216
several thermophilic and hyperthermophilic Thermotogae genomes (Doerrler et al. 2013). This 217
suggests that the function provided by DedA is either not needed by these organisms, or is being 218
provided by analogous integral membrane proteins, or that their DedA homologs are too 219
divergent to be detected by sequence similarity searches.
220 221
Nucleic acids: a challenge to keep the strands together 222
11 High temperatures denature double stranded DNA and secondary structures of RNA. This
223
presents a problem for thermophiles, and for hyperthermophiles in particular. These organisms 224
must maintain their chromosomes in an orderly state for both efficient packaging as well as 225
coordinated gene expression. Therefore, to survive the damaging effects of high temperature 226
thermophiles need to either continuously repair their damaged DNA or protect it from damage in 227
the first place. For example, the archaeon Pyrococcus abyssi has a highly efficient DNA repair 228
system that continuously repairs temperature-induced DNA damage (Jolivet et al. 2003). Very 229
high levels of homologous recombination are observed in hyperthermophilic Thermotoga spp.
230
where the ratio of nucleotide changes introduced by recombination relative to point mutation 231
(r/m) is in the range 24-100 for genomes originating from geographically distant sites (Nesbø et 232
al. 2006a; Nesbø et al. 2014). This in the upper range of values reported in a comparison of r/m 233
across a large sample of mostly mesophilic Bacteria and Archaea (0.02 – 64), where values 234
above 10 were interpreted as very high (Vos and Didelot 2009). The high level of recombination 235
may be explained by the need for DNA repair in thermophiles (Johnston et al. 2014). This 236
hypothesis is supported by observations of high levels of recombination and repair in other 237
hyperthermophilic microorganisms, such as Pyrococcus furiosus (DiRuggiero et al. 1997), 238
Sulfolobus islandicus (Whitaker et al. 2005), and Persephonella (Mino et al. 2013).
239
Protection of DNA is known to occur via multiple unrelated mechanisms. Primarily, 240
thermophiles safeguard their DNA with thermostable proteins analogous to eukaryotic histones.
241
For example, in the archaeon Thermococcus kodakaraensis HpkA and HpkB dramatically 242
increase the melting temperature of a given DNA sequence upon binding, with HpkB being able 243
to raise the melting temperature of poly(dA-dT) DNA by > 20°C (Higashibata et al. 1999), 244
suggesting that these proteins play a major role in the stabilization of Thermococcus 245
12 kodakaraensis chromosomes. In the bacterium T. maritima the histone-like protein HU stabilizes 246
and protects the DNA (Mukherjee et al. 2008).
247
Thermophiles can also use polyamine compounds to stabilize their DNA and RNA, as 248
well as many other cellular components. Multivalent polyamine compounds such as putrescine, 249
spermidine, and spermine, or their acetylated forms, compact histone-bound DNA in 250
Thermococcus kodakaraensis, stabilizing it at temperatures as high as 90°C (Higashibata et al.
251
2000). In Thermotoga species the polyamines caldopentamine and caldohexamine increase in 252
concentration with increased temperature, suggesting a role in thermal response and thermal 253
adaptation (Zellner and Kneifel 1993). Indeed caldopentamine and caldohexamine, as well as five 254
other long linear polyamines found in Thermus thermophilus, have been shown to stabilize 255
double-stranded DNA at high temperature, with a greater stabilizing effect by polyamines with a 256
larger number of amino nitrogen atoms (Terui et al. 2005).
257
Thirdly, unique RNA modifications can confer thermostability in thermophiles 258
(McCloskey et al. 2001). For example, modifications from adenosine to 2’-O-methyladenosine or 259
from guanosine to N2,2’-O-dimethylguanosine in the tRNAs are often growth temperature- 260
specific, even among closely related lineages (McCloskey et al. 2001).
261
Lastly, thermal adaptation may be achieved via reverse gyrase-mediated DNA 262
supercoiling. Reverse gyrase is a protein found almost exclusively in hyperthermophiles and, 263
importantly, it is a gene carried by all known hyperthermophiles (Brochier-Armanet and Forterre 264
2006; Forterre 2002; Lulchev and Klostermeier 2014). While deletion of the reverse gyrase gene 265
from Thermococcus kodakaraensis results in slower growth at high temperatures (90°C), it does 266
not abolish its growth, suggesting that this enzyme is not essential for hyperthermophilic growth 267
as was once thought (Atomi et al. 2004). However, since the T. kodakaraensis mutant lacking 268
13 reverse gyrase grew poorly at 90°C, and unlike the wild-type strain, could not grow above 90°C 269
(Atomi et al. 2004), this enzyme is still considered to be a critical adaptation for optimal growth 270
at high temperatures (Brochier-Armanet and Forterre 2006). Although reverse gyrase catalyzes 271
ATP-dependent positive supercoiling of DNA in vitro, its function in vivo remains unknown. The 272
increased heat protection provided by this enzyme may be linked to a role in the DNA damage 273
response, possibly through recruitment to lesions (Lulchev and Klostermeier 2014; Perugino et al.
274
2009). Interestingly, cultivated hyperthermophilic species from both the Thermotogae and the 275
Aquificae have acquired their reverse gyrase genes from Archaea by lateral gene transfer (LGT), 276
suggesting that hyperthermophily may have been acquired by Bacteria from Archaea (Brochier- 277
Armanet and Forterre 2006; Forterre et al. 2000).
278
While some of these adaptations for nucleic acid stabilization have only been found in 279
thermophiles (e.g., reverse gyrase (Forterre 2002), certain RNA modifications (McCloskey et al.
280
2001) and thermostable histones (Higashibata et al. 1999)), others are found in mesophiles as well.
281
For instance, the same polyamines found in Thermotoga are also found in mesophilic microalgae 282
(Nishibori et al. 2009). Hence, transition between thermophily and mesophily may only require a 283
re-purposing of certain cellular constituents, rather than removing or acquiring them.
284
In addition to cellular components interacting with nucleic acids for stabilization, the 285
composition of some nucleic acids appears adapted to the thermophilic lifestyle of the host 286
organism. The extra hydrogen bond in G:C nucleotide pairs was long thought to play a part in 287
optimal growth temperature. While genome-wide G+C content does not correlate with optimal 288
growth temperature (Galtier and Lobry 1997; Hurst and Merchant 2001; Zeldovich et al. 2007), 289
the G+C content of some structural RNA encoding genes does. For example, the G+C content of 290
secondary structures of rRNA and tRNA molecules, specifically in the stem structures, increases 291
14 with optimal growth temperature (Galtier and Lobry 1997; Kimura et al. 2013; Zhaxybayeva et al.
292
2009). As a result, the GC content variation of the 16S rRNA gene can be used as a proxy for 293
studying temperature adaptation within the Thermotogae. For example, the temperature optimum 294
for uncultured members of the phylum was predicted by establishing a correlation between the 295
16S rRNA gene distances and optimal growth temperature of 33 Thermotogae isolates (Dahle et 296
al. 2011). Additionally, inference of the ancestral states of the 16S rRNA gene that gave rise to 30 297
Thermotogae isolates allowed Green et al. (2013) to hypothesize that the thermotolerant 298
Thermotogae lineages are under directional selection and that transition from high to low optimal 299
growth temperature is easier to achieve.
300 301
Compatible solutes: the power of redundancy 302
Compatible solutes are organic compounds that are accumulated by cells under stressful 303
conditions such as osmotic stress and heat stress (Santos et al. 2011). These compounds, 304
particularly polyamines, are known to stabilize nucleic acids in thermophilic cells (see above).
305
Moreover, in the bacterium Calderobacterium hydrogenophilum polyamine compounds stabilize 306
the 70S initiation complex of ribosomes (Mikulik and Anderova 1994). Many temperature studies 307
in the Thermotogae have focused on the accumulation of these organic compounds and 308
polyamines and the elucidation of their biosynthetic pathways in T. maritima and the more 309
moderate thermophile Petrotoga miotherma (Jorge et al. 2007; Oshima et al. 2011; Rodionova et 310
al. 2013; Rodrigues et al. 2009; Zellner and Kneifel 1993). Several compatible solutes have so far 311
only been found in thermophiles including di-myo-inositol phosphate, mannosyl-di-myo-inositol 312
phosphate, mannosylglyceramide, and diglycerol phosphate (Borges et al. 2010; Gonçalves et al.
313
2012) and novel thermophilic solutes continue to be identified (Jorge et al. 2007; Rodrigues et al.
314
15 2009). However, while these compounds are thermophile-specific and may represent thermophile- 315
specific adaptations, they are not the only compatible solutes used to deal with heat stress. When 316
the ability to synthesize di-myo-inositol phosphate was removed from Thermococcus 317
kodakarensis by deleting a key synthesis gene, the growth of this archaeon was unaffected, and 318
aspartate accumulated as an alternative compatible solute (Borges et al. 2010). In the 319
Thermotogae multiple solutes accumulate under stress conditions (Jorge et al. 2007; Rodrigues et 320
al. 2009). This suggests that although the role compatible solutes play in thermal protection is not 321
fully understood, there is functional redundancy among the solutes.
322 323
Protein dynamics and turnover; assistance from chaperones and proteases 324
Chaperones are large protein complexes that assist the proper folding and re-folding of proteins.
325
The chaperonins represent an extensively studied subclass of chaperones with a stacked double-ring 326
structure (Large et al. 2009). Distribution of the chaperone families varies across Bacteria and 327
Archaea, and some chaperones are considered indispensable (Large et al. 2009). For example, some 328
chaperonins help fold new polypeptides, as well as re-fold and rescue proteins that have been 329
inactivated due to stress (Techtmann and Robb 2010). A major stressor that triggers chaperone- 330
mediated protein repair is heat shock, which has resulted in many chaperones being named heat 331
shock proteins (HSP) (Large et al. 2009). By preventing inactivation and aggregation of proteins at 332
high temperatures, this ubiquitous system is thought to be especially important in thermophiles, 333
which employ chaperones in both unstressed and heat-stressed states (Pysz et al. 2004). Thus, while 334
these proteins are part of high temperature response in mesophiles, their constitutive expression in 335
thermophiles may be part of their temperature adaptation. For example, the predicted chaperone 336
TM1083 in T. maritima is thought to stabilize the DNA gyrase enzyme at temperatures near optimal 337
16 growth (Canaves 2004). Moreover, the molecular chaperone trigger factor (TM0694) from T.
338
maritima strongly binds model proteins and decreases their folding rate, while these activities are 339
much weaker in the homologous trigger factor from the psychrophile Pseudoalteromonas 340
haloplanktis, which instead shows increased prolyl isomerization (Godin-Roulling et al. 2014).
341
However, it should be noted that chaperones, although always highly expressed in thermophiles, are 342
part of their high temperature response as well. For instance, examination of the T. maritima 343
proteome at four temperatures spanning its growth range revealed higher relative abundance of 344
chaperones at supra-optimal temperatures (Wang et al. 2012).
345
Proteases are also part of the heat shock response in mesophilic organisms (Richter et al.
346
2010). A key distinction between well-studied bacterial mesophiles and the hyperthermophile T.
347
maritima is the lack of regulation in T. maritima of most of its proteases in response to 348
temperature stress (Conners et al. 2006). This may be explained by an absence of major 349
regulators of the mesophilic proteolytic response (i.e., rpoH or ctsR homologs) in the T. maritima 350
genome (Conners et al. 2006; Pysz et al. 2004). Perhaps this bacterium gains a survival 351
advantage from constitutive expression of most proteases. A similarity search revealed an 352
absence of detectable rpoH and ctsR homologs in 38 Thermotogae, including the thermophilic K.
353
olearia and the mesophilic M. prima, suggesting that any regulation of protease expression in the 354
Thermotogae involves different genes than those used by other Bacteria and Archaea.
355
356
Thermal adaptation at the protein level 357
Although chaperones aid in proper folding and maintenance of proteins under high temperature 358
conditions, proteins from thermophilic organisms are themselves adapted to high temperature.
359
This adaptation is required to maintain activity at temperatures that would denature mesophilic 360
17 homologs and is found at all levels of protein structure, from primary through quaternary. Protein 361
thermostability is also not uniform across the proteome and depends on its functional role:
362
proteins either having catalytic activity or regulating other catalytic proteins appear to be under 363
greater selection to be temperature adapted than proteins involved in, for example, core 364
transcriptional or translational processes (Gu and Hilser 2009).
365
While there are many examples of specific thermostabilizing characteristics and 366
interactions at each of the four levels of globular protein structure (reviewed by Imanaka (2011) 367
and Li et al. (2005)), there is no universal property that confers thermostability. Rather, it is the 368
combination of factors at all levels of structure that grants high temperature activity in globular 369
proteins. Increased thermostability is often due to slight differences in sequence and structure, and 370
thermophilic and mesophilic counterparts are typically very similar proteins (Taylor and Vaisman 371
2010). Below we briefly overview known pathways to temperature adaptation in globular proteins.
372
Protein primary structure is the amino acid sequence of the polypeptide chain. Ultimately, 373
the properties and sequence of the amino acids determine the final higher level structures of the 374
protein. One characteristic associated with thermostable proteins is enrichment of amino acids that 375
contribute to a strong hydrophobic core. Larger aliphatic amino acids with more branches are 376
favored at positions that fill cavities, which may ultimately strengthen the protein through 377
increased hydrophobic interactions (Clark et al. 2004). Taylor and Vaisman (2010), however, 378
found that it is only a moderately good indicator of protein thermostability.
379
Comparisons of amino acid composition of thermophilic and mesophilic proteins have 380
revealed several trends at the global proteome level. The observed excess of charged (D,E,K,R) 381
versus polar (N,Q,S,T) amino acids in soluble proteins from hyperthermophiles, known as the 382
CvP bias (Cambillau and Claverie 2000; Gao and Wang 2012; Holder et al. 2013; Suhre and 383
18 Claverie 2003), may reflect larger importance of ionic interactions between charged amino acids 384
over hydrogen-bond interactions for retaining protein structure as temperature increases 385
(Cambillau and Claverie 2000). Additionally, a systematic evaluation of all possible subsets of 386
amino acids revealed that the total fraction of the amino acids IVYWREL in a proteome most 387
strongly correlates with optimal growth temperature (Zeldovich et al. 2007).
388
The CvP and IVYWREL biases have been explored thoroughly in the Thermotogae where 389
both indices show strong linear correlations with optimal growth temperature (Zhaxybayeva et al.
390
2009). Specifically, the distribution of CvP values was unimodal for each of the Thermotogae 391
proteomes, arguing against the hypothesis that thermophily is a recently acquired trait of the 392
Thermotogae. Moreover, calculation of CvP values from estimated ancestral Thermotogae 393
sequences suggested that the ancestral Thermotogae proteome belonged to organisms with an 394
optimal growth temperature of ≈84.5°C, higher than that of any characterized extant Thermotogae 395
bacterium (Zhaxybayeva et al. 2009). While the average CvP value for most of the thermophilic 396
Thermotogae lineages was above 10.62, the mesophilic M. prima proteome has an average CvP 397
value of 8.96 (Zhaxybayeva et al. 2012). Also this genome has a unimodal CvP distribution, 398
suggesting it has maintained a mesophilic lifestyle for a long time. An exception to the trend is 399
observed in the P. lettingae genome, which has an average CvP value of 8.42 (Zhaxybayeva et al.
400
2009), but an optimal growth temperature of 65°C. However, P. lettingae-like 16S rRNA genes 401
and genomic DNA have been recovered from environments with temperatures < 65°C (e.g., 40- 402
50°C, (Nesbø et al. 2010; Nobu et al. 2014)), suggesting that these bacteria often live at 403
temperatures below the optimal growth temperature of the cultivated isolate.
404
Protein secondary structure describes the local folding of polypeptide sequences. This 405
includes regular structures like α-helices and β-sheets, or irregular structures like β-turns, coils 406
19 and loops. These are formed primarily by hydrogen bond interactions between the backbone and 407
side chain elements of the amino acids. In addition to having secondary structures that facilitate 408
tighter packing and rigidity at the tertiary level, thermophilic proteins tend to have secondary 409
structures that are more stabilized than their mesophilic counterparts (Facchiano et al. 1998; Koga 410
et al. 2008; Prakash and Jaiswal 2010). For example, thermostable proteins have been reported to 411
have a larger fraction of their amino acid residues arranged in α-helices than mesophilic proteins 412
do (Prakash and Jaiswal 2010).
413
Protein tertiary structure is the arrangement of a folded polypeptide chain in three- 414
dimensional space. This is achieved by disulfide bridges, electrostatic interactions within the 415
polypeptide chain, and hydrophobic interactions and hydrogen bonding within the chain as well as 416
between the peptides and solvent. Thermophilic proteins tend to have conformations that are more 417
rigid and more tightly packed, with reduced entropy of unfolding and conformational strain 418
compared to their mesophilic counterparts (Li et al. 2005). The strongest contributors to 419
thermostability are increased ion pairs on the protein surface combined with a more strongly 420
hydrophobic interior (Taylor and Vaisman 2010). In agreement with this, additional salt bridges 421
on the surface of the enzyme diguanylate cyclase from T. maritima accounted for its greater 422
thermostability compared to the same enzyme found in the mesophiles Pseudomonas aeruginosa, 423
Marinobacter aquaeolei and Geobacter sulfurreducens (Deepthi et al. 2014). Additionally, the 424
glutamate dehydrogenase enzymes of the hyperthermophilic bacterium T. maritima and 425
hyperthermophilic archaeon P. furiosus have smaller hydrophobic accessible surface area (ASA) 426
and greater charged ASA than the glutamate dehydrogenase from the mesophilic bacterium 427
Clostridium symbiosum (Knapp et al. 1997). Since few other structural differences were found 428
20 between the thermophilic and mesophilic enzymes, this tighter packing is thought to contribute to 429
the thermal stability of the proteins.
430
Protein quaternary structure is the arrangement of multiple folded polypeptide chains into 431
a multimeric complex. In globular proteins this level of structure is formed and maintained by 432
many of the same forces that contribute to the tertiary structure of a protein, but between 433
polypeptide chains rather than within them. These forces include disulfide bridges, electrostatic 434
interactions, hydrophobic interactions and hydrogen bonding. In thermostable proteins, greater 435
numbers of these interactions, or stronger interactions over weaker ones, are favored (Li et al.
436
2005).
437
One additional way of achieving greater protein stability is to increase the number of 438
subunits. For example, the malate dehydrogenase (MDH) enzyme, which is usually a dimer in 439
mesophiles, is a tetramer in the thermophilic bacterium Chloroflexus aurantiacus (Bjørk et al.
440
2003). The additional dimer-dimer interface of the tetrameric MDH is hypothesized to provide 441
thermal stability due to the higher number of inter-polypeptide interactions compared to the 442
mesophilic dimers. To test this hypothesis, Bjørk et al. (2003) introduced a disulfide bridge that 443
would strengthen dimer-dimer interaction further, and found that the new enzyme had a melting 444
temperature 15°C higher than the wild-type enzyme. In addition, removing excess negative charge 445
at the dimer-dimer interface by replacing a glutamate residue with either glutamine or lysine 446
resulted in an increase of apparent melting temperature by ~ 24°C (Bjørk et al. 2004).
447 448
Tolerating new temperatures: Is it possible to modify just a few proteins?
449
As discussed above, adaptation to a high optimal growth temperature is achieved differently by 450
Bacteria and Archaea, by one species than another, and even by one protein than another within 451
21 the same organism. Given that all of these factors combine in unique ways to permit growth 452
within a specific temperature range, how could a shift in permissive temperature range be 453
accomplished? While some of these strategies are universal to thermophiles and mesophiles, such 454
as utilization of chaperones and compatible solutes, others, like shifting of membrane properties, 455
would have to be radically altered to accommodate large changes in temperature range.
456
Changing a few key proteins may have global stabilizing effects on the whole cell. For 457
instance, some of the proteins whose stability appears most affected by thermal adaptation are 458
involved in production of compatible solutes that stabilize other proteins (Gu and Hilser 2009).
459
Such changes would reduce the need to modify the stability of all components of the proteome. It 460
may also be possible to lower the maximal growth temperature of an organism through changes to 461
a single protein (Endo et al. 2006). By replacing the chromosomal copy of groEL chaperonin in 462
Bacillus subtilis 168 (growth range from 11 to 52°C) with a psychrophilic groEL from 463
Pseudoalteromonas sp. PS1M3 (growth range from 4 to 30°C), Endo and colleagues noted a 2°C 464
reduction in the maximal growth temperature of the newly constructed B. subtilis strain. Similarly, 465
the heterologous expression of a small heat shock protein from Caenorhabditis elegans, enabled E.
466
coli cells to grow at temperatures up to 50°C (and survive heat shock at 58°C for 1/2h) extending 467
its growth range by 3.5°C (Ezemaduka et al. 2014). While these changes do not constitute true 468
shifts in growth temperature range or changes to optimal growth temperature, these studies 469
suggest that changes to a single key protein (involved both in temperature adaptation and 470
response) could extend or narrow the temperature range at which an organism is able to grow by a 471
few degrees. Accumulation of several such mutations could eventually lead to a more substantial 472
shift in growth range. Some of these mutations may be advantageous at lower temperatures, while 473
others may be loss-of-function mutations, where abilities to function at higher temperatures are 474
22 lost for proteins in individuals living in an environment with temperatures at the lower end of 475
their original growth range. Under the latter scenario, change in the growth temperature range 476
might not be a result of selection, but rather a product of random genetic drift or genetic 477
hitchhiking with another, unrelated trait selected for in the new environment.
478 479
Role of Lateral Gene Transfer in Temperature Adaptation: Acquisition of Already 480
‘Adapted’ Genes 481
Lateral gene transfer (LGT) is a major force in prokaryotic evolution, allowing rapid adaptation to 482
changes in the environment by acquiring clusters of genes or single genes that confer a selective 483
advantage (Boucher et al. 2003; Zhaxybayeva and Doolittle 2011) and LGT has been implicated 484
in adaptation to extreme environments including high temperatures (see for example Omelchenko 485
et al. (2005)). Genes encoding proteins linked to adaptation to higher or lower growth 486
temperatures have been laterally exchanged (reviewed in Boucher et al. 2003). Reverse gyrase is 487
a classic example of lateral transfer of a single gene that is thought to have been crucial for 488
evolutionary adaptation to high temperatures by hyperthermophilic Bacteria (Brochier-Armanet 489
and Forterre 2006; Forterre 2002). Phylogenetic analyses suggest two ancient acquisitions of this 490
gene by bacterial lineages from Archaea, followed by secondary transfer events among Bacteria 491
(Brochier-Armanet and Forterre 2006).
492
Similarly, the compatible solute di-myo-inositol phosphate is thought to be important for 493
heat tolerance in thermophiles and hyperthermophiles (Borges et al. 2010). Two key genes 494
involved in the synthesis of this compound (inositol-1-phosphate cytidylyltransferase and di-myo- 495
inositol phosphate phosphate synthase) are suggested to have been laterally transferred from an 496
archaeal lineage to hyperthermophilic marine Thermotoga species, while in other lineages the two 497
23 genes are predicted to have fused before being exchanged among several bacterial and archaeal 498
lineages (Gonçalves et al. 2012).
499
Reverse gyrase and the myo-inositol pathway genes are just two examples of a large 500
number of genes transferred into the Thermotogae. When the genome of T. maritima MSB8 was 501
first sequenced (Nelson et al. 1999), 24% of its open reading frames (ORFs) showed greatest 502
similarity to sequences from Archaea, suggesting that these genes have been acquired from these 503
distantly related organisms that inhabit the same environment. Comparative genomic analyses of 504
additional Thermotogae genomes have confirmed an influx of genes from Archaea (albeit the total 505
number dropped to 10-11% of the ORFs, due to increased number of bacterial homologs in 506
GenBank) and an even larger fraction of Firmicutes genes in these genomes (Mongodin et al.
507
2005; Nesbø et al. 2009; Zhaxybayeva et al. 2009; Zhaxybayeva et al. 2012). Phylogenetic 508
analysis of all the ORFs in the M. prima genome suggests this lineage has undergone extensive 509
gene exchange with diverse mesophilic lineages, and that LGT has aided its transition from a 510
thermophilic to a mesophilic lifestyle (Zhaxybayeva et al. 2012). Thus, as a major force that has 511
shaped the genomes of the Thermotogae, LGT may have also been important for the acquisition 512
and development of the temperature ranges of the various Thermotogae lineages. Most of the 513
acquired genes in Thermotogae (including M. prima) are involved in carbohydrate metabolism 514
(Mongodin et al. 2005; Nesbø et al. 2009; Zhaxybayeva et al. 2009; Zhaxybayeva et al. 2012).
515
However, M. prima has additionally acquired genes involved in signal transduction mechanisms, 516
secondary metabolite biosynthesis, and amino acid transport and metabolism (Zhaxybayeva et al.
517
2012), suggesting the potential importance of genes from these functional categories for life at 518
lower temperatures.
519 520
24 Transition to mesophily in Kosmotoga and Mesotoga
521
The discovery of the mesophilic Thermotogae lineage (Mesotoga) raised the possibility that 522
(hyper)thermophily was not ancestral to the phylum. However, as discussed above, the amino acid 523
composition (CvP bias and IVYWREL amino acids frequency) of the reconstructed ancestral 524
Thermotogae proteome suggests that the ancestral Thermotogae was a hyperthermophile 525
(Zhaxybayeva et al. 2009), and that the transition to mesophily in the Thermotogae phylum is 526
secondary. Moreover, ancestral sequence reconstruction of myo-inositol-3-phosphate synthase 527
enzymes in the Thermotoga genus also suggests that the ancestor of this hyperthermophilic 528
lineage grew optimally at temperatures higher than those of extant species (Butzin et al. 2013).
529
The G+C content of ribosomal RNA, which correlates with optimal growth temperature, also 530
suggests that the reconstructed 16S rRNA of the ancestor of all Thermotogae belonged to a 531
thermophile (Green et al. 2013; Zhaxybayeva et al. 2009).
532
So far, the genus Mesotoga is the only strictly mesophilic Thermotogae, with optimal 533
growth occurring between 37 and 45°C (Ben Hania et al. 2013; Nesbø et al. 2012). Initially 534
Mesotoga spp. were only detected using molecular tools such as community 16S rRNA PCR and 535
metagenome analyses (Nesbø et al. 2006b). Mesotoga prima was the first described isolate of the 536
genus (Nesbø et al. 2012), which now includes another validated species, Mesotoga infera, (Ben 537
Hania et al. 2013), one yet to be validated, Mesotoga sp. PhosAc3 (Ben Hania et al. 2011), and 538
several isolates with ongoing genome sequencing projects (Benson et al. 2014; Reddy et al.
539
2014). The 2.97 Mb genome of M. prima is considerably larger than any previously sequenced 540
Thermotogae genome, which range between 1.86 and 2.30 Mb (Zhaxybayeva et al. 2012). This 541
larger size is due to both higher numbers of protein-coding genes and larger intergenic regions. A 542
unimodal distribution of CvP values of M. prima's proteome, with a mean value in the 543
25 mesophilic range, indicate that native M. prima proteins have also changed in response to its 544
evolved mesophilic lifestyle (Zhaxybayeva et al. 2012).
545
Analysis of additional Thermotogae shows that the variation in size may be related to 546
optimal growth temperature: thermophiles have more streamlined genomes, with little intergenic 547
space and a higher number of genes per transcription unit, while mesophiles have larger 548
intergenic spaces and higher gene redundancy (Latif et al. 2013; Zhaxybayeva et al. 2012). This 549
finding holds true for lineages outside of the Thermotogae, as examination of 1155 prokaryotes 550
demonstrates (Sabath et al. 2013). However, the observed correlation in Thermotogae needs to 551
be untangled from effects of phylogenetic history (Zhaxybayeva et al. 2012).
552
The closest relative of the Mesotoga lineage is the thermophilic lineage Kosmotoga (Fig.
553
3). Members of this genus have been found in hydrothermal sediments (L'Haridon et al. 2014;
554
Nunoura et al. 2010) and oil production fluids (DiPippo et al. 2009; Feng et al. 2010). Like other 555
thermophilic Thermotogae, the Kosmotoga are anaerobic chemoorganotrophs able to ferment 556
carbohydrates and peptides (Nunoura et al. 2010) and to produce molecular hydrogen (DiPippo 557
et al. 2009; Feng et al. 2010). The first isolated bacterium of this genus was Kosmotoga olearia 558
(DiPippo et al. 2009). K. olearia grows optimally at 65°C and has a reported growth range of 20- 559
80°C (DiPippo et al. 2009). Not only is this bacterium capable of growing at an unusually low 560
temperature for a thermophile, but to our knowledge it represents the widest reported bacterial 561
temperature growth range to date.
562
The ability of Kosmotoga to grow at such an extraordinary gamut of temperatures is 563
intriguing for two reasons. First, it must maintain protein activity and membrane integrity. Every 564
living organism has adapted to do this at a certain temperature range, but how these requirements 565
can be maintained over a 60°C range is unknown. What evolutionary mechanisms would maintain 566
26 a 60°C growth interval in Kosmotoga? Perhaps this lineage continues to experience environments 567
with more variable temperatures or, alternatively, the wide growth range may be a result of 568
selection on another trait. Second, as discussed above, this ability of tolerating a wide range of 569
temperature conditions, may have facilitated the transition of Mesotoga from thermophily to 570
mesophily, because the capacity to grow at lower temperatures presumably already existed in 571
Mesotoga's ancestors.
572
As a result Kosmotoga and Mesotoga offer a unique model system for studying both 573
immediate temperature responses and long-term temperature adaptation. Specifically, K. olearia’s 574
exceptionally wide growth range allows examination of temperature responses under both 575
mesothermic and thermic conditions in the same cell-line. For example, analysis of K. olearia’s 576
transcriptome at different growth temperatures promises to shed light into the role of specific 577
processes, functions, genes or proteins in thermoadaptation. Since K. olearia’s closest relative is a 578
mesophile with a narrower growth range, comparative genomic, transcriptomic and proteomic 579
analyses promise to reveal how Kosmotoga's temperature responses may eventually lead to 580
temperature adaptation. If we assume that Mesotoga and Kosmotoga’s common ancestor was a 581
thermophile, possibly with a wide growth range, then the Mesotoga lineage lost its ability to grow 582
at high temperatures, while Kosmotoga has either kept or expanded its growth range. For 583
Mesotoga we have speculated that reduction of its growth temperature range may have happened 584
as the lineage got 'trapped' in an oil reservoir that cooled down (Nesbø et al. 2006b; Zhaxybayeva 585
et al. 2012) and therefore may be a result of loss-of-function mutations and genetic drift.
586
The existence of several additional Thermotogae lineages likely thriving in mesothermic 587
environments (Nesbø et al. 2010) opens opportunities to study the evolutionary processes in 588
lineages that have adapted to lower temperatures independently. These novel lineages can be 589
27 accessed through metagenomic studies or through further cultivation efforts. Taken together, 590
future genomic, transcriptomic and proteomic studies of temperature responses and adaptations 591
in Kosmotoga, Mesotoga, and other Thermotogae will help decipher how shifts in temperature 592
range and optimum are accomplished.
593 594
Acknowledgements 595
This work is supported by an NSERC Alexander Graham Bell Canada Graduate Scholarship 596
CGS-M to S.M.J.P., by a Norwegian Research Council of Norway award (project no.
597
180444/V40) to C.L.N., and by a Simons Investigator award from the Simons Foundation, 598
Dartmouth’s Walter and Constance Burke Research Initiation Award and Dean of Faculty start- 599
up funds to O.Z.
600 601
References 602
603
Achenbach-Richter, L., Gupta, R., Stetter, K.O., and Woese, C.R. 1987. Were the original 604
eubacteria thermophiles? Syst. Appl. Microbiol. 9(1-2): 34-39.
605
Adl, S.M., Simpson, A.G.B., Lane, C.E., Lukeš Julius, Bass, D., Bowser, S.S. et al. 2012. The 606
revised classification of eukaryotes. J. Eukaryot. Microbiol. 59(5): 429-493.
607
Akanuma, S., Nakajima, Y., Yokobori, S., Kimura, M., Nemoto, N., Mase, T., et al. 2013.
608
Experimental evidence for the thermophilicity of ancestral life. Proc. Natl. Acad. Sci. U. S. A.
609
110(27): 11067-11072.
610
Arti, D., Park, J., Jung, T.Y., Song, H., Jang, M., Han, N.S. et al. 2012. Structural analysis of α- 611
L-Arabinofuranosidase from Thermotoga maritima reveals characteristics for thermostability and 612
substrate specificity. J.Microbiol.Biotechnol. 22(12): 1724-1730.
613
Atomi, H., Matsumi, R., and Imanaka, T. 2004. Reverse gyrase is not a prerequisite for 614
hyperthermophilic life. J. Bacteriol. 186(14): 4829-4833.
615
28 Ben Hania, W., Ghodbane, R., Postec, A., Brochier-Armanet, C., Hamdi, M., Fardeau, M. et al.
616
2011. Cultivation of the first mesophilic representative (mesotoga) within the order 617
Thermotogales. Syst. Appl. Microbiol. 34(8): 581-585.
618
Ben Hania, W., Postec, A., Aüllo, T., Ranchou-Peyruse, A., Erauso, G., Brochier-Armanet, et al.
619
2013. Mesotoga infera sp. nov., a mesophilic member of the order Thermotogales, isolated from 620
an underground gas storage aquifer. Int. J. Syst. Evol. Microbiol. 63: 3003-3008.
621
Benson, D.A., Clark, K., Karsch-Mizrachi, I., Lipman, D.J., Ostell, J., and Sayers, E.W. 2014.
622
GenBank. Nucleic Acids Res. 43: D30-D35.
623
Bhandari, V., and Gupta, R.S. 2014. Molecular signatures for the phylum (class) Thermotogae 624
and a proposal for its division into three orders (Thermotogales, Kosmotogales ord. nov. and 625
Petrotogales ord. nov.) containing four families (Thermotogaceae, Fervidobacteriaceae fam.
626
nov., Kosmotogaceae fam. nov. and Petrotogaceae fam. nov.) and a new 627
genus Pseudothermotoga gen. nov. with five new combinations. Antonie van Leeuwenhoek 628
105(1): 143-168.
629
Bjørk, A., Dalhus, B., Mantzilas, D., Sirevåg, R., and Eijsink, V.G.H. 2004. Large improvement 630
in the thermal stability of a tetrameric malate dehydrogenase by single point mutations at the 631
dimer-dimer interface. J. Mol. Biol. 341(5): 1215-1226.
632
Bjørk, A., Dalhus, B., Mantzilas, D., Eijsink, V.G.H., and Sirevåg, R. 2003. Stabilization of a 633
tetrameric malate dehydrogenase by introduction of a disulfide bridge at the dimer-dimer 634
interface. J. Mol. Biol. 334(4): 811-821.
635
Borges, N., Matsumi, R., Imanaka, T., Atomi, H., and Santos, H. 2010. Thermococcus 636
kodakarensis mutants deficient in di-myo-inositol phosphate use aspartate to cope with heat 637
stress. J. Bacteriol. 192(1): 191-197.
638
Boucher, N., and Noll, K.M. 2011. Ligands of thermophilic ABC transporters encoded in a 639
newly sequenced genomic region of Thermotoga maritima MSB8 screened by differential 640
scanning fluorimetry. Appl. Environ. Microbiol. 77(18): 6395-6399.
641
Boucher, Y., Douady, C.J., Papke, R.T., Walsh, D.A., Boudreau, M.E.R., Nesbø, C.L., et al.
642
2003. Lateral gene transfer and the origins of prokaryotaic groups. Annu. Rev. Genet. 37: 283- 643
328.
644
Boussau, B., Blanquart, S., Necsulea, A., Lartillot, N., and Gouy, M. 2008. Parallel adaptations 645
to high temperatures in the Archaean eon. Nature 456(7224): 942-945.
646
Brochier-Armanet, C., and Forterre, P. 2006. Widespread distribution of archaeal reverse gyrase 647
in thermophilic bacteria suggests a complex history of vertical inheritance and lateral gene 648
transfers. Archaea 2(2): 83-93.
649
29 Butzin, N.C., Lapierre, P., Green, A.G., Swithers, K.S., Gogarten, J.P., and Noll, K.M. 2013.
650
Reconstructed Ancestral Myo-Inositol-3-Phosphate Synthases Indicate That Ancestors of the 651
Thermococcales and Thermotoga Species Were More Thermophilic than Their Descendants.
652
PloS one 8(12): e84300.
653
Cambillau, C., and Claverie, J. 2000. Structural and genomic correlates of hyperthermostability.
654
The Journal of biological chemistry 275(42): 32383-32386.
655
Canaves, J.M. 2004. Predicted role for the Archease protein family based on structural and 656
sequence analysis of TM1083 and MTH1598, two proteins structurally characterized through 657
structural genomics efforts. Proteins 56(1): 19-27.
658
Cappelletti, M., Zannoni, D., Postec, A., and Ollivier, B. 2014. Members of the order 659
Thermotogales: From microbiology to hydrogen production. In Microbial BioEnergy: Hydrogen 660
Production. Edited by D. Zannoni and R. De Philippis. Springer Netherlands, Dordrecht. pp. 197- 661
224.
662
Carballeira, N.M., Reyes, M., Sostre, A., Huang, H., Verhagen, M.F., and Adams, M.W. 1997.
663
Unusual fatty acid compositions of the hyperthermophilic archaeon Pyrococcus furiosus and the 664
bacterium Thermotoga maritima. J. Bacteriol. 179(8): 2766-2768.
665
Chang, E.L. 1994. Unusual thermal stability of liposomes made from bipolar tetraether lipids.
666
Biochem. Biophys. Res. Commun. 202(2): 673-679.
667
Chhabra, S.R., Shockley, K.R., Ward, D.E., and Kelly, R.M. 2002. Regulation of endo-acting 668
glycosyl hydrolases in the hyperthermophilic bacterium Thermotoga maritima grown on glucan- 669
and mannan-based polysaccharides. Appl. Environ. Microbiol. 68: 545-554.
670
Chhabra, S.R., Shockley, K.R., Conners, S.B., Scott, K.L., Wolfinger, R.D., and Kelly, R.M.
671
2003. Carbohydrate-induced differential gene expression patterns in the hyperthermophilic 672
bacterium Thermotoga maritima. J. Biol. Chem. 278(9): 7540-7552.
673
Clark, A.T., McCrary, B.S., Edmondson, S.P., and Shriver, J.W. 2004. Thermodynamics of core 674
hydrophobicity and packing in the hyperthermophile proteins Sac7d and Sso7d. Biochemistry 675
43: 2840-2853.
676
Cobucci-Ponzano, B., Zorzetti, C., Strazzulli, A., Carillo, S., Bedini, E., Corsaro, M.M. et al.
677
2011. A novel α-D-galactosynthase from Thermotoga maritima converts β-D-galactopyranosyl 678
azide to α-galacto-oligosaccharides. Glycobiology 21(4): 448-456.
679
Comfort, D.A., Bobrov, K.S., Ivanen, D.R., Shabalin, K.A., Harris, J.M., Kulminskaya, A.A. et 680
al. 2007. Biochemical analysis of Thermotoga maritima GH36 α-galactosidase (TmGalA) 681
confirms the mechanistic commonality of clan GH-D glycoside hydrolases. Biochemistry 682
46(11): 3319-3330.
683
30 Conners, S.B., Mongodin, E.F., Johnson, M.R., Montero, C.I., Nelson, K.E., and Kelly, R.M.
684
2006. Microbial biochemistry, physiology, and biotechnology of hyperthermophilic Thermotoga 685
species. FEMS Microbiol. Rev. 30(6): 872-905.
686
Cuneo, M.J., Beese, L.S., and Hellinga, H.W. 2009. Structural analysis of semi-specific 687
oligosaccharide recognition by a cellulose-binding protein of Thermotoga maritima reveals 688
adaptations for functional diversification of the oligopeptide periplasmic binding protein fold. J.
689
Biol. Chem. 284(48): 33217-33223.
690
Dagan, T., Roettger, M., Bryant, D., and Martin, W. 2010. Genome networks root the tree of life 691
between prokaryotic domains. Genome Biology and Evolution 2: 379-392.
692
Dahle, H., Hannisdal, B., Steinsbu, B.O., Ommedal, H., Einen, J., Jensen, et al. 2011. Evolution 693
of temperature optimum in Thermotogaceae and the prediction of trait values of uncultured 694
organisms. Extremophiles 15(4): 509-516.
695
Damsté, J.S.S., Rijpstra, W.I.C., Hopmans, E.C., Schouten, S., Balk, M., and Stams, A.J.M.
696
2007. Structural characterization of diabolic acid-based tetraester, tetraether and mixed 697
ether/ester, membrane-spanning lipids of bacteria from the order Thermotogales. Arch.
698
Microbiol. 188(6): 629-641.
699
de Mendoza, D. 2014. Temperature sensing by membranes. Annu. Rev. Microbiol. 68: 101-116.
700
Deepthi, A., Liew, C.W., Liang, Z., Swaminathan, K., and Lescar, J. 2014. Structure of a 701
diguanylate cyclase from Thermotoga maritima: Insights into activation, feedback inhibition and 702
thermostability. PloS one 9(10): e110912.
703
DiDonato, M., Deacon, A.M., Klock, H.E., McMullan, D., and Lesley, S.A. 2004. A scaleable 704
and integrated crystallization pipeline applied to mining the Thermotoga maritima proteome. J.
705
Struct. Funct. Genomics 5: 133-146.
706
DiPippo, J.L., Nesbø, C.L., Dahle, H., Doolittle, W.F., Birkland, N., and Noll, K.M. 2009.
707
Kosmotoga olearia gen. nov., sp. nov., a thermophilic, anaerobic heterotroph isolated from an oil 708
production fluid. Int. J. Syst. Evol. Microbiol. 59: 2991-3000.
709
D'Ippolito, G., Dipasquale, L., and Fontana, A. 2014. Recycling of carbon dioxide and acetate as 710
lactic acid by the hydrogen-producing bacterium Thermotoga neapolitana. ChemSusChem 7:
711
2678-2683.
712
DiRuggiero, J., Santangelo, N., Nackerdien, Z., Ravel, J., and Robb, F.T. 1997. Repair of 713
extensive ionizing-radiation DNA damage at 95 degrees C in the hyperthermophilic archaeon 714
Pyrococcus furiosus. J. Bacteriol. 179(14): 4643-4645.
715
Doerrler, W.T., Sikdar, R., Kumar, S., and Boughner, L.A. 2013. New functions for the ancient 716
DedA membrane protein family. J. Bacteriol. 195(1): 3-11.
717
31 Endo, A., Sasaki, M., Maruyama, A., and Kurusu, Y. 2006. Temperature adaptation of Bacillus 718
subtilis by chromosomal groEL replacement. Biosci. Biotechnol. Biochem. 70(10): 2357-2362.
719
Ezemaduka, A.N., Yu, J., Shi, X., Zhang, K., Yin, C., Fu, X. et al. 2014. A small heat shock 720
protein enables Escherichia coli to grow at a lethal temperature of 50°C conceivably by 721
maintaining cell envelope integrity. J. Bacteriol. 196(11): 2004-2011.
722
Facchiano, A., Colonna, G., and Ragone, R. 1998. Helix-stabilizing factors and stabilization of 723
thermophilic proteins: an X-ray based study. Protein Eng. 11(9): 753-760.
724
Feng, Y., Cheng, L., Zhang, X., Li, X., Deng, Y., and Zhang, H. 2010. Thermococcoides 725
shengliensis gen. nov., sp. nov., a new member of the order Thermotogales isolated from oil- 726
production fluid. Int. J. Syst. Evol. Microbiol. 60: 932-937.
727
Forterre, P., Bouthier De La Tour C, Philippe, H., and Duguet, M. 2000. Reverse gyrase from 728
hyperthermophiles: probable transfer of a thermoadaptation trait from Archaea to Bacteria.
729
Trends in genetics : TIG 16(4): 152-154.
730
Forterre, P. 2002. A hot story from comparative genomics: reverse gyrase is the only 731
hyperthermophile-specific protein. Trends in genetics : TIG 18(5): 236-237.
732
Frock, A.D., Gray, S.R., and Kelly, R.M. 2012. Hyperthermophilic Thermotoga species differ 733
with respect to specific carbohydrate transporters and glycoside hydrolases. Appl. Environ.
734
Microbiol. 78(6): 1978-1986.
735
Galtier, N., and Lobry, J.R. 1997. Relationships between genomic G+C content, RNA secondary 736
structures, and optimal growth temperature in prokaryotes. J. Mol. Evol. 44(6): 632-636.
737
Gao, J., and Wang, W. 2012. Analysis of structural requirements for thermo-adaptation from 738
orthologs in microbial genomes. Annals of Microbiology 62(4): 1635-1641.
739
Ghimire-Rijal, S., Lu, X., Myles, D.A., and Cuneo, M.J. 2014. Duplication of genes in an ATP- 740
binding cassette transport system increases dynamic range while maintaining ligand specificity.
741
J. Biol. Chem. 289(43): 30090-30100.
742
Godin-Roulling, A., Schmidpeter, P.A.M., Schmid, F.X., and Feller, G. 2014. Functional 743
adaptations of the bacterial chaperone trigger factor to extreme environmental temperatures.
744
Environ. Microbiol. in press.
745
Gogarten, J.P., Kibak, H., Dittrich, P., Taiz, L., Bowman, E.J., Bowman, et al. 1989. Evolution 746
of the vacuolar H+-ATPase: implications for the origin of eukaryotes. Proc. Natl. Acad. Sci. U.
747
S. A. 86(September): 6661-6665.
748
Gonçalves, L.G., Borges, N., Serra, F., Fernandes, P.L., Dopazo, H., and Santos, H. 2012.
749
Evolution of the biosynthesis of di-myo-inositol phosphate, a marker of adaptation to hot marine 750
environments. Environ. Microbiol. 14(3): 691-701.
751